Feb. 5, 1952
NOTES
207-208', was obtained. A mixed m.8. of the picrate with the free base was depressed to 150-170 A n d . Calcd. for Cl&IuNbOa: N, 17.81. Found: N, 17.97. 2-Methyl-5,6-trimethylenepydol4 (2-Methyl6,7-dihydrocyclopentapyrimidol4).-The general procedure was used with 10 ml. of ethanol, 0.47 g. of acetamidine hydrochloride and 0.78 g. of 2-carbethoxycyclopentanone' (b.p. 99-100.5' (10 mm.)). The reaction time was 12 hours a t 25' plus 3.5 hours at reflux temperature. The extraction solvent was ethyl acetate. On cooling, the extract deposited long colorless needles. These when collected, washed with a little ethyl acetate, and dried, weighed 180 mg., m.p. 208-212'. From the Ntrate by partial evaporation 54 mg. more material, of m.p. 207-212', was obtained, making the total yield 234 mg. (31%). The material sublimed easily at 2 mm. (125' bath), but this did not improve the purity. Repeated recrystallization of the material from ethyl acetate raised the m.p. to 214.5215.5'. Analysis indicated that the compound was still not entirely pure, but it gave a picrate of satisfactory purity. Anal. Cafcd. for CsHioNtO: C, 63.98; H, 6.71; N, 18.66. Found: C, 64.91; H. 7.10; N, 19.36. 4-Hydroxy-2-methyl-5,6-trimethylenepyrimidinium Picrate.-The crude picrate, prepared as described for the above quinazoline, was obtained in 51% yield, m.p. 178.5182.5'. After recrystallization from n-butanol it melted at 184.5-185'. Anal. Calcd. for C1,HlaNsOs: C, 44.33; H, 3.45; N, 18.47. Found: C, 44.11; H, 3.84; N, 18.29. 2-Carbethoxy-4-methoxycyclohexanone .-Hydrogenation&of commercial 4-methoxyphenol gave a cis-trans diastereomeric mixture of 4-methoxycyclohexanol, b.p. 98107' (15 mm.), in 89% yield. The product was oxidized according to the procedure of H e l f d but to obtain the reported yield it was necessary to reduce the proportion of dichromate to a slight excess. 4-Methoxycyclohexanone was obtained in 50-55% yield as a colorless liquid, b.p. 82-83' (13 mm.). A sample was converted to the 2,4-dinitrophenylhydrazone, m.p. 147148' (reported' 150'). The method of Cook and Lawrence6 for the preparation of 2-carbethoxyformyl-4-methoxycyclohexanonewas modified as follows: The mixture of 4-methoxycyclohexanone (11.4 g.) and diethyl oxalate (14.7 9.) was added to the solution of sodium ethoxide in absolute ethanol (distilled from sodium). The reaction was conducted a t 0-5' under dry nitrogen. After 2-3 hours stirring at O', the mixture was left overnight (with stirring), and after chilling to 0' was acidified and extracted. The decarbonylation procedures was modified by adding 100 mg. of powdered glass and 10 mg. of powdered iron to the entire batch of crude diketo-ester (a viscous oil), heated in a Claisen flask at 190'. The vigorous evolution of carbon monoxide was complete in 20 minutes, and 4.42 g. (25%, based on methoxycyclohexanone) of 2-carbethoxy4methoxycyclohexanone, b.p. 128-135' (14 mm.), was obtained. On the uroduct was obtained as a colorless - - redistillation .- -. . liquid (3.69 g.), b.p.-139-141' (18 mm.) (reported* 131133" (10 mm.)), nmD 1.4783, Dzo 1.102. M% 51.40 (theor. 49.50). The product gave a 2,4dinitrophenylhydrazoneof m.p. 125-128' dec., reported8 129-131' dec.
.
843
In earlier experiments the free base was isolated from the reaction mixture as colorless crystals, m.p. 124-237'. The yield was only 12% and purification was difficult. 5-Carbethoxymethy1-2,6-dimethylpyrimidol-4..-The general procedure was used with acetamidine hydrochloride (0.47 g.), 10 ml. of ethanol and 1.08 g. of diethyl acetosuc&ate' (b.p. 103.6106.5" (1 mm.), #OD 1.4348, M ~ D 52.22 (theor. 52.81). The reaction time was three days at 25'. The extraction solvent was ethyl acetate. The colorless crystals which separated from the extract were collected, washed (ethyl acetate) and dried, weight 159 mg. The fdtrate on vacuum distillation to half-volume and dilution 1: 1 with benzene gave 24 mg. more, making the total yield 182 mg. (18%), m.p. 177-179.5". After recrystallization of the product from ethyl acetate-benzene (1 :1) the m.p. was 181.5-182'. Anal. Calcd. for Cld&N20,: C, 57.13; H, 6.71; N, 13.33. Found: C, 57.80; H, 6.59; N, 13.39.
5-Carbethoxymethyl-2-hy&oxymethyl-6-methylp~dol4.-The general procedure was used with 1.10 g. of hydroxyacetamidine hydrochloride* (m.p. 147-148.5'), 22 ml. of ethanol and 2.16 g. of diethyl acetosuc~inate.~(The amidine HCl was predissolved in half of the ethanol a t the b.p.) The reaction time was 24 hours at 25' followed by 3 hours refluxing. The extraction solvent was 40 ml. of benzene (refluxed 20 minutes). The hot filtered extract on cooling deposited colorless needles which were collected, washed (benzene), and dried, giving 330 mg. (15%) of product, m.p. 137-142'. Repeated recrystallization of the product from benzene raised the m.p. to 143-144.5'. Anal. Calcd. for CloHtrNzO,: C, 53.09; H , 6.24; N, 12.39. Found: C, 52.48; H,6.03; N, 11.61. On heating with hydrazine hydrate the ester yields a colorless crystalline hydrazide, which can be recrystallized from ethanol and melts a t about 225' (rapid heating), but which has not yet been purified sufficiently for analysis. (9) H. Adkins, N. Isbell and B. Wojcik, ref. 3, p. 262.
DEPARTMENT OF CHEMISTRY UNIVERSITY OF TORONTO RECEIVED SEPTEMBER 21, 1951 TORONTO, CANADA
Anion Exchange Studies. IV.112 Cobalt and Nickel in Hydrochloric Acid Solutions BY GEORGEE. MOOREAND KURTA. KRAUS
The anion exchange behavior of nickel(I1) and cobalt(I1) in 0.5 to 12 M hydrochloric acid was studied by measuring their elution constants E = d A / V , where d is the distance (cm.) an absorption band moves when V (cc.) of eluent have been passed through a column of A (cm.2)cross-sectional area. Glass columns of the practically colorless resin Dowex-1 (a "strong base" quaternary amine polystyrene-divinylbenzene resin) were used, whose 6-Methoxy-2-methyl-5,6,7,&tetrahydroquinazolinol-4 Hy- cross-sectional areas were 0.023 cm.2. Since both drochloride.-The general procedure was used with 0.47 g. nickel and cobalt are colored (the latter has a of 2-carbethoxy4methoxycyclohexanone. The reaction characteristic blue color on the resin) the band The extraction solvent was positions (and hence E ) were determined visually. time was three days at 25'. absolute ethanol. The hot filtered extract was cooled to 0'. Dry hydrogen chloride was passed into the solution for Approximately 1 mg. of Co and Ni were used per ten minutes. The colorless crystalline precipitate was col- experiment. In addition a few experiments were lected, washed with ethanol and dried, giving 0.62 g. (540/o) carried out with tracer cobalt (CosO)and d deterof product, m.p. 217-220' dec. After recrystallization mined with the automatic scanner previously of the product from absolute ethanol the m.p. was 221-222' described.a dec. The results for cobalt are summarized in Fig. 1. Anal. Calcd. for CloHl~CIN~OI:C, 52.06; H, 6.55; Eco rapidly decreases from approximately 2.5 N. 12.15. Found: C, 52.24; H , 6.94;. N,. 12.15. in less than 3 M HC1 to ca. 0.02 in 9 M HCl and (4) R. P. Linstead and E. M. Meade, ibid., p. 119. (5) E. M. Van Duzee and H. Adkins, TEIS JOURNAL, 67, 147 (1935). (6) L. Helfer, Helo. Chim. Acto, 7, 050 (1924). (7) C. S. Marvel and W. L. Walton, J . Or#. Chem., 7,88 (1942). (8) J. W. Cook and C. A. Lawrence, J. Chcm. Soc., 68 (1988).
(1) This document is based on work performed for the Atomic Enerm Commission at the Oak Rfdne National Laboratorv. (2Y-Revious paper: K. A. K r a u s k d G. E. Moore, T H IJOURNAL, ~ 78. 2900 (1951). i3) K. A. k a u 8 and G. E. Moore, ibid., 78, 9 (1951).
VOl. 74
NOTES
844
then increases again to ca. 0.055 in 12 N HCl. earlier conclusions. Additional evidence for the reasonably general applicability of this rule comes E N i on the other hand was found to be ca. 2.5 throughout the whole acid range. The difference from the observation that gallium (probably as in the adsorption behavior of the two elements is GaC14-) is adsorbed quantitatively similarly to iron sufficient to permit excellent separation. Since (FeC14-).e One might thus tentatively assume as discussed earlier3 E = l/(i 0)where i is the that FeCl4- and CoCla- are also approximately fractional interstitial space and D the distribution equally strongly adsorbed. On this basis the fraccoefficient (amount per ml.. of resin/amount per ml. tion of cobalt as CoCla- can be estimated to be conof solutions), small values of E indicate good ad- siderably less than 1% even in 9 M BC1, the adsorption. Furthermore, E will reach a maximum sorption maximum. value (Emax = l/i) when D becomes zero. Using Since in concentrated HC1 a negatively charged sodium and potassium tracers and assuming that complex of cobalt almost certainly exists in contheir ions are not adsorbed E m a x = ca. 2.5 was siderable c o f i c e n t r a t i ~ none ~ ~ ~can ~ ~ conclude that found for these columns. Hence there is negligible it is not strongly adsorbed. Since this ion probably adsorption of nickel in the range 0.5 to 12 M HC1 has a charge of minus two (C0Cl4-)~1*this relatively and of cobalt in the range 0.5 to 3 M. poor adsorption in concentrated HCl is in general agreement with the earlier conclusion that in this medium only singly negatively charged complexes are extremely strongly adsorbed. The non-adsorbability of Ni(I1) by the resin suggests very strongly that negatively charged I 4 I complexes of this element are not formed in appreciable concentration even in concentrated hydrochloric acid.
+
f i-
I
r\
(6) K. A. Kraus, F. Nelson and G. W. Gtnith, unpublished results. (7) M. Bobtelsky and E. 9. Spirglw, J . Chcm. Sac., 142 (1849). (8) N. V. Sidgwick, “Thr Chemicel Elements and Their Compounds,” Vot. XI, Clntendon Prer, Oxford, 1950. (9) According to recent paper elwfxalnigration work cobalt begihs to migrate toward the anode In ca. 8 M R.A. Kmur and 0.W. Smith, unpublished results; fot technique see T H f s JOWRNAL, ?a, 4529 (1850).
Ha;
c
v
0 s RIDGE NATIONAL LABORATORY OAK RIDGE, TENN. RECSIVEDJULY 18, 1951
Ethyl a,a-Dibromoy,y,y-trifluoroacetoacetate BY WILLIAML. MOSEY
AIthough the preparation of ethyl y,y,y-trifluoroacetoacetate was described some time ago,l little seems to be known of ita a-halogenated derivatives. The preparation and properties of ethyl a-chloro-y,y,y-trifluoroacetoacetate are described by Hill, Towne and Dickey.2 The bromination of ethyl y,y,y-trifluoroacetoacetateis mentioned by Heme and Mencher,*but no details are given. Ethyl a,a-dibromo-y,y,y-tuoroacetoacetate has been prepared by the bromination of ethyl y, y! y-trifluoroacetoacetate in the preaence of pyndine, using (a) bromine itself, and (b) pyridine perbromide hydr~bromide.~Both methods gave essentially the same rather low yield of product. In view of the facile chlorination of ethyl y,?,ytrifluoroacetoacetate,2 perhaps the simpler direct treatment of the ester with bromine would be effective, but as no further work on this compound is contemplated, our data are presented now. The considerable fore-run obtained in the distillations probably represents products formed by cleavage of the ester. Ethyl a,a-dibromo-r,y,y-trifluoroacetoacetate is a colorless, mobile liquid, with a slight, but not unpleasant, odor (in contrast to (1) F. Swarts, Bwll. Sci. Acad. Ror. Balg., 151 13, 679 (1950). M.Hill, E. B. Towne and J. B. Dickey, THISJOURNAL, 72, 3289 (1950). (3) A. L. Henne and L. Mencher, Abstracts of Papers, 118th Meeting A. C.S , Chicago, Ill., Sept. 3-8, 1950, p. 1OL. (4) C . Djerassi and C. R Scholz, THISJ O U R N A L , 70, 417 (1948) (2) H.
